Two of the present limitations associated with X-ray CT perfusion are X-ray dose and the use of iodinated contrast materials. In the work proposed here we intend to test the hypotheses that CT perfusion can be done using one tenth of the conventional X-ray dose per time frame using iodinated contrast material. Our group has recently developed a unique data acquisition and reconstruction strategy called HYPR (HighlY constrained back PRojection) that exploits the redundancy of information in a temporal series of images. Basically, a composite image is formed from all acquired projections. The composite image is then used to constrain the backprojection of the data acquired in individual time frames. The method has provided acquisition acceleration factors as high as several hundred in MR angiography but acceleration is reduced in other applications for which greater numbers of projections are required to characterize the time frames. The degree of acceleration depends on the sparsity of the data in the composite image. We propose to extend the HYPR method to X-ray CT where the use of a reduced number of projections per time frame results in reduced x-ray dose. The SNR characteristics of HYPR are such that the SNR generated by all projections in the temporal series feeds back into the SNR of the individual time frames. This represents a completely new SNR behavior in which SNR of individual time frames is proportional to the square root of overall scan time rather than the square root of the frame time. The investigation involving iodinated contrast material will consist of simulations, phantom studies, animal studies and clinical evaluations. We also propose to investigate to perform simulations, phantom studies and animal studies to investigate the feasibility of performing perfusion studies at current or possibly reduced dose levels using gadolinium contrast materials.